pcpg_rna <- readRDS("Data/pcpg_with_metadata_and_qc.RDS")
pcpg_rna$cell_subtype <- recode(pcpg_rna$cell_subtype, "Sustentacular cells" = "SCLCs")
Supplementary Figure UMAPs
genes_of_interest <- c(
"PNMT",
"LEF1",
"EPAS1",
"RET",
"IRX4",
"EGLN3"
)
metadata_of_interest_pcpg <- c(
"Processing_batch",
"Patient",
"Cell_Type",
"pseudohypoxic_or_kinase_signalling")
metadata_of_interest <- c(
"Cell_Type",
"Processing_batch",
"Patient",
"pseudohypoxic_or_kinase_signalling",
"tumor_subtype",
"phase") # TODO: add in annotation columns for individual genotype true or false for tumor only
# FIGURE SX --------------------------------------------------------
# panel A: batch UMAP
# panel B: Normal Chromaffin cells colored by sample UMAP
# panel C: PGL1 and PGL3 tumor cells colored by sample
# TODO make UMAPs where all cells greyed out except a single genotype, color by sample
# make table with metadata, UMAP coords and normalised, scaled gene expression
pcpg_rna_table <- getPlotData(pcpg_rna, genes_of_interest)
Genotypes <- c("MAX", "NF1", "MAML3", "SDHD")
genotype_columns <- c()
for (i in seq_along(Genotypes)){
genotype <- Genotypes[i]
varname <- paste0(genotype, "_highlighted")
print(varname)
pcpg_rna_table[[varname]] <- if_else(pcpg_rna_table$Genotype == genotype &
pcpg_rna_table$Cell_Type == "Tumour",
pcpg_rna_table$Sample,
"Wildtype")
genotype_columns[i] <- varname
}
## [1] "MAX_highlighted"
## [1] "NF1_highlighted"
## [1] "MAML3_highlighted"
## [1] "SDHD_highlighted"
# make metadata columns which will be used to colour the umaps
pcpg_rna_table$Chromaffin_cells_highlighted <- if_else(pcpg_rna_table$Sample %in% c("E240", "E243") &
pcpg_rna_table$Cell_Type == "Chromaffin cells",
pcpg_rna_table$Sample,
"Wildtype")
pcpg_rna_table$PGL1_PGL3_highlighted <- if_else(pcpg_rna_table$Sample %in% c("P018-PGL1", "P018-PGL3") &
pcpg_rna_table$Cell_Type == "Tumour",
pcpg_rna_table$Sample,
"Wildtype")
pcpg_rna_table$Processing_batch <- factor(pcpg_rna_table$Processing_batch,
levels = as.character(1:7))
# umaps
batch.cols <- setNames(brewer.pal("Dark2", n=7),
as.character(1:7))
c1c2_cols <- c("C1" = "red", "C2" = "blue", "Normal" = "lightgrey")
samples <- unique(pcpg_rna_table$Sample)
sample_cols <- pals::polychrome(n = 33)[2:34]
names(sample_cols) <- c("Wildtype", samples)
md_columns_fig1 <- c("Processing_batch", "Chromaffin_cells_highlighted", "PGL1_PGL3_highlighted")
colorpals_fig1 <- list(batch.cols, sample_cols, sample_cols)
fig1_plot_list <- map2(.x = md_columns_fig1,
.y = colorpals_fig1,
.f = function(x, y){ggUMAPplot(data = pcpg_rna_table,
group.by=x,
colorpal=y)})
fig1_umaps <- wrap_plots(fig1_plot_list, ncol = 3, guides = "collect") &
theme(legend.position = 'bottom')
fig1_umaps
# ggsave(plot = fig1_umaps,
# width = 300, height = 300, units = "mm",
# filename = "Figures/rebuttal_fig1_umaps.pdf")
#
# ggsave(plot = fig1_umaps, dpi=600,
# width = 300, height = 300, units = "mm",
# filename = "Figures/rebuttal_fig1_umaps.png")
# FIGURE SX ------------------------------------------------------------------------------------
# Panel A: C1 and C2 subtypes
# Panel C-E: samples within specific genotypes
md_columns_fig2 <- c("pseudohypoxic_or_kinase_signalling", genotype_columns)
colorpals_fig2 <- list(c1c2_cols, sample_cols, sample_cols, sample_cols, sample_cols)
fig2_plot_list <- map2(.x = md_columns_fig2,
.y = colorpals_fig2,
.f = function(x, y){ggUMAPplot(data = pcpg_rna_table,
group.by=x,
colorpal=y)})
fig2_umaps <- wrap_plots(fig2_plot_list, ncol = 3, guides = "collect") &
theme(legend.position = 'bottom')
fig2_umaps
# ggsave(plot = fig2_umaps,
# width = 300, height = 300, units = "mm",
# filename = "Figures/rebuttal_fig2_umaps.pdf")
# ggsave(plot = fig2_umaps, dpi=300,
# width = 300, height = 300, units = "mm",
# filename = "Figures/rebuttal_fig2_umaps.png")
# FIGURE SX: ------------------------------------------------------------------------------------
# Panel A-F: Harmony batch-corrected UMAPS
# Panel G-L: featureplots
pcpg_rna_harmonized <- readRDS("Results/pcpg_harmonized_patient.RDS")
# make table with metadata, UMAP coords and normalised, scaled gene expression
harmony_table <- getPlotData(pcpg_rna_harmonized, genes_of_interest)
harmony_table$Processing_batch <- factor(harmony_table$Processing_batch,
levels = as.character(1:7))
harmony_table$tumor_subtype_highlighted <- if_else(harmony_table$Cell_Type == "Tumour",
harmony_table$tumor_subtype, "Normal",)
md_columns_fig3 <- c("Cell_Type", "Patient", "Processing_batch", "pseudohypoxic_or_kinase_signalling", "tumor_subtype_highlighted", "Phase")
subtype_genotype_cols2 <- c(subtype_genotype_cols, "Normal" = "lightgrey")
colorpals_fig3 <- list(cell.cols, "auto", batch.cols, c1c2_cols, subtype_genotype_cols2, "auto")
fig3_umap_list <- map2(.x = md_columns_fig3,
.y = colorpals_fig3,
.f = function(x, y){ggUMAPplot(data = harmony_table,
group.by=x,
colorpal=y)})
# featureplots
fig3_feature_plot_list <- map(.x = genes_of_interest,
.f = function(x){ggFeaturePlot(harmony_table, x)})
fig3 <- wrap_plots(c(fig3_umap_list,fig3_feature_plot_list), ncol = 3, guides = 'collect')
fig3
# check featureplot gene order:
# wrap_plots(fig3_feature_plot_list, ncol = 3)
# ggsave(plot = fig3,
# width = 600, height = 400, units = "mm", dpi = 600,
# filename = "Figures/fig3_harmony_featureplots.png")
# ggsave(plot = fig3,
# width = 600, height = 400, units = "mm",
# filename = "Figures/fig3_harmony_featureplots.pdf")
# FIGURE SX: ------------------------------------------------------------------------------------
# Endothelial cell subclustering plots
# get the endothelial cells UMAPs
endothelial_table <- getPlotData(pcpg_rna, genes_of_interest, reduction = "Endothelial.cells")
cell.subtype.cols <- setNames(pals::polychrome(n=27), unique(pcpg_rna$cell_subtype))
md_columns_endothelial <- c("cell_subtype", "Processing_batch", "Patient")
endothelial_table$Processing_batch <- factor(endothelial_table$Processing_batch,
levels = as.character(1:7))
colorpals_endothelial <- list(cell.subtype.cols[unique(endothelial_table$cell_subtype)],
batch.cols, "auto")
# endothelial umaps
umap_list_endothelial <- map2(.x = md_columns_endothelial,
.y = colorpals_endothelial,
.f = function(x, y){ggUMAPplot(data=endothelial_table,
group.by=x,
colorpal=y)})
endothelial <- wrap_plots(umap_list_endothelial, ncol = 3, guides = 'collect')
endothelial
# ggsave(plot = endothelial,
# width = 400, height = 200, units = "mm", dpi = 600,
# filename = "Figures/endothelial_umaps_rebuttal.png")
# ggsave(plot = endothelial,
# width = 400, height = 200, units = "mm",
# filename = "Figures/endothelial_umaps_rebuttal.pdf")
